Early bone ingrowth and minimal resorption ensure rigid fixation in uncemented total knee replacements. Trabecular titanium–aluminum–vanadium (Ti6Al4V) and hydroxyapatite (HA)-coated vacuum-plasma-sprayed (VPS) titanium with varying porosities and HA-coating thicknesses, have been developed to enhance fixation, though bone cellular response remains largely unknown. This study evaluated osteoblast responses to trabecular Ti6Al4V and three VPS titanium surfaces with 20%–40% or 30%–70% porosity and HA coatings of 60, 80, or 90 µm. Human primary osteoblasts were seeded and cultured for 29 days, to assess seeding efficiency, viability, metabolic activity, alkaline phosphatase activity, and the effect of osteoblast-released factors in conditioned medium during the last 4 days of culture on osteoclast formation. VPS-HA groups were first compared individually; when no differences were found, data were pooled for comparison with the trabecular group. Osteoblast seeding efficiency, viability, metabolic activity, and alkaline phosphatase activity were similar between VPS-HA surfaces. Moreover, osteoblast-conditioned medium did not affect osteoclast formation. Osteoblast seeding efficiency and viability were similar between the pooled VPS-HA and trabecular surface. Compared to the pooled VPS-HA, the trabecular surface increased osteoblast metabolic (1.5-2.7-fold) and alkaline phosphatase activity (6.5-15.2-fold). Osteoblast-conditioned medium reduced osteoclast formation (2.1-3.4-fold) on trabecular compared to the pooled VPS-HA surface. In conclusion, these findings show that VPS-HA surfaces with various porosities and HA-coating thicknesses similarly affect osteoblast and osteoclast responses, while trabecular surfaces enhance osteoblast responsiveness and inhibit osteoclast formation. These results might help to further improve early stability and reduce long-term loosening risk in uncemented knee replacements.